Chapter 1
Philip II, king of the Macedonians from 359 to 336 BC, campaigned several times into Thrace, the lands to the north and east of his kingdom, and over the period of his reign he succeeded in conquering large areas of it. He failed to conquer several cities on the Propontis (the Sea of Marmara), where Perinthos and Byzantion failed to succumb to his attacks, but in the Thracian interior he was able to gain control of the land as far as the Haemos Mountains (i.e. the Balkan Mountains), and was able to establish a couple of Macedonian colonies and several forts in that territory to maintain his control.1
The Macedonian control which was thus established over much of Thrace was less than firm, and it may better be characterized as superficial. It appears to have consisted of a few major garrisons at such places as Philippopolis and Kabyle, where colonists were established; these were linked by smaller posts or forts which were spread along the connecting roads, though as a means of maintaining control this was unlikely to have been successful for very long.2 A number of the local Thracian chiefs had survived by submitting formally to Macedonian authority; several of these were members of the Odrysian royal family which had ruled in the area for the previous two centuries, still active. Such submissions were only good for the lifetime of Philip and his interlocutor, however; further trouble would ensue when one of them died.
The Thracian mode of succession was to divide the kingdom among all males with a claim, which made for weakness, especially in the face of the new Macedonian aggressiveness. The Greek cities on the coasts of the Aegean, the Propontis and on the Black Sea were largely under Philip’s control or were friendly towards him, but this was in part the result of Thracian hostility, and holding both the cities and Thrace in the same regime would be difficult. Such arrangements as Philip made for the government of the conquests relied more on the threat of retaliation if trouble arose than on conciliation and protection. No governor for the conquered region was appointed by Philip, as far as we know; Alexander did appoint one when he was about to set off on his campaign in Asia Minor, which emphasized the explicitly personal nature of Philip’s conquests.
Beyond the Haemos Mountains Philip had made diplomatic contacts with other tribes and rulers, and had conducted some raids, especially along the Black Sea coastal area, and even across the Danube, but no permanent conquests had been made.3 His northern boundary was thus the Haemos range. On the other hand, those raids into the north had been clear messages to the inhabitants north of the mountains that conquest might well come at any time, and Philip’s diplomacy was always aggressive, seeking submissions. To the kings and chiefs of the land between the Haemos Mountains and the Danube River, it will have been obvious that they were likely to be the next on Philip’s target list, though, as it happened, the next were the Persians. The whole scheme of conquests and colonization in the north had a highly provisional look to it; if Philip was serious about holding this land, he would need to deploy time and troops, and to find good reasons to do so – but he was after other prey.
Under the circumstances it is a little surprising that the murder of Philip in 336 was not followed by an instant Thracian rebellion. There were still a number of members of the old Thracian royal houses of the Odrysai alive and active who could have led such a rising, but their recent submissions will have restrained them.
In the event the trouble when it came was a war with the Triballi, who were located north of the Haemos range and west of the conquered region, between those mountains and the Danube. This people had refrained from becoming involved in Philip’s conquests, but had watched carefully for the opportunity to snatch some advantage from the wreckage he was making of other areas; in 339, they had intercepted a baggage train of loot and captives gathered by Philip in a raid on the Skythians – ‘20,000’ captives and ‘20,000’ horses and other still more portable loot, according to the sources, which probably reflected Macedonian annoyance. Philip had not reacted, even though he himself was wounded in the fight.4 Possibly his wound immobilized him for a time, but he had several competent generals who could have administered punishment. No doubt the Triballi were marked in his mind for future punishment. No doubt also this success will have whetted Triballian appetites.
The new Macedonian king, Alexander III, having swiftly dealt with the usual Illyrian raids which always followed on the death of a Macedonian king, and having dealt similarly with stirrings of troubles in Greece, turned to attack the Triballi. He was opposed at the pass through the Haemos and broke through; the Triballians evacuated their families to an island in the Danube, and stood to fight in defence of their lands and people. Alexander approached in battle formation, and sent ships along the Danube as a threat to the island. The Triballi were again defeated. Both battles are said to have cost only fifty Macedonian casualties between them, but about 5,000 Triballi died; these figures may even be correct, at least approximately.
The second defeat brought the submission of the Triballian king, Syrmos, and a peace agreement was made. Arrian reports that other Thracian tribes also now submitted; these were presumably Thracians who had originally submitted to Philip; renewing their submissions to Alexander would establish Macedonian control for the rest of Alexander’s own lifetime.
Alexander captured the major Danubian crossing point at the modern town of Zimnicea (the later Roman town of Novae, or Ad Novas), during that Triballian war, though the town was burnt in the process. Beyond the river he now encountered his first Galatians. Like the Triballi, when Philip was campaigning in Thrace the Galatians had been watching events south of the river with interest, and now joined in the peace talks, sending an embassy to make contact with Alexander.5
This embassy implies that the envoys represented a fairly wellorganized polity. Just as Syrmos negotiated on behalf of the kingdom of the Triballi, so the envoys of the Galatians were speaking on behalf of an organized state. It is customary to speak of these peoples, Galatians and Triballi, as tribes, but it is clear that Syrmos had a powerful authority over his people, and so must be counted as their king. The Galatian embassy was sent by some similar tribal-wide authority, a king or, more likely, a ruling aristocracy, probably from the nearest organized Galatian community, and therefore the one most concerned. The envoys had come from a considerable distance, for Strabo claims they represented ‘Kelts who lived about the Adriatic’,6 suggesting that he did not really know where they came from, though it was clearly from a considerable distance, and from north of the Danube.
The Galatian state was, in fact, presumably that of the Scordisci, located in the Banat region. This was a new-minted tribe which was confected from groups detached from the Taurisci and Boii to the north. The reference to the Adriatic suggests that some of these Galatians at least were also, or were later, on the Adriatic coast. This means, however, that the Galatians who were living by the Danube were part of the same state as the envoys. They had been in that area for perhaps seventy years, which is plenty of time for the invaders to organize themselves as a Galatian tribe; the rulers were the leaders of the invasion and their genetic successors. The name Scordisci may not yet have been adopted, but may be used here for convenience.
There had evidently been time for the Galatians north of the Danube to report on Alexander’s Triballian campaign – they will no doubt have noted Philip’s campaign as well – and then for the envoys to travel from the Banat to the neighbourhood of Zimnicea in time to attend the peace conference. Alexander’s campaign had evidently taken some time, for he had to arrange for his ships to sail from the Propontis (or even from the Aegean) to the Danube and then up the river, which would have taken perhaps two or three weeks to accomplish. The Kelts’ embassy could have been organized and dispatched to arrive well in time to watch the burning of the bridge-town after the final Triballian defeat. There would be no reason for them to be on the spot earlier; it was only as a response to Alexander’s campaign against the Triballi that they came so far from the centre of Galatian power. Had Alexander, as Philip before him, continued his work of conquest in the Balkans as far as the Danube, he would have clearly met and probably fought the Scordisci fairly soon. For the moment he was some distance away, and his destruction of the town and the river crossing implied that he had no intention of campaigning further; but from the town to the Banat was not very far.
The Galatians asked for Alexander’s friendship, or so Arrian puts it, but then adds that ‘they exchanged pledges’. This was therefore an agreement between equals, not a submission of the Galatians to Alexander. The two sides were agreeing not to go to war with each other, so that both sides could get on with their own particular plans without fear of an attack from the rear by the other. This clearly suited Alexander, since his intention was to take up his father’s plan and campaign against the Persian Empire; it obviously also suited the Galatians, though we can only guess at their plans from what happened later. It is a confirmation of the Galatians’ assumption of political equality with Alexander that their reply when he asked them what the Galatians feared most, fairly obviously expecting to be nominated himself, was to instance their fear of the sky falling on them, or the earth opening up to swallow them, or the sea overwhelming them. That is, they feared the gods, not men. There was also a threat buried in that reply, as well as a more obvious assertion of equality. The agreement they had made was, that is, to be regarded as only a temporary arrangement, which might be denounced by either side if necessary; in the way of such agreements it would last only until one party to it died – it therefore expired in 323, on the death of Alexander.7
As it was, of course, both of the Macedonian kings had their eyes fixed on richer game. Philip had begun an invasion of the Persian Empire in the year before he died, and, as soon as Alexander had dealt with his various European and Macedonian problems in the same superficial way as his father in Thrace, he set off to conquer the east. The Galatians, having made a first contact with Alexander and the Macedonians, renewed that contact when the king returned from his Indian adventure to Babylon, being one of numerous sets of envoys who were no doubt principally anxious to discover what Alexander’s plans were.8 It has been suggested that the agreement they had made at the Danube meeting was that the Galatians would protect the Macedonian frontier on the Danube side while Alexander was off campaigning,9 but there is no actual evidence for this other than that the peace they had made endured during Alexander’s lifetime in accordance with normal diplomatic practice, and indeed if they made a peace agreement (‘exchanged pledges’) there would be no need for anything more detailed. Certainly the Galatians did not interfere when there was a major Thracian revolt in 331, nor again after a major Macedonian defeat north of the Danube near the Black Sea coast in 325. The best interpretation for the envoys’ visit to Babylon in 325/324 is that they were there to congratulate the king, like almost everybody else in the crowd of envoys, and to try to discover his plans for the Balkans; a renewal of their former peace agreement was perhaps their main intention as a means of fending off a possible attack. They could fully appreciate the difference made by Alexander’s extensive conquests in terms of his ability to exercise his power. We may note that, like all the other envoys, they had been fully aware of Alexander’s progress, and duly understood that his return to Babylon suggested a further campaign in the Mediterranean area. The publication of Alexander’s plans after his death can only have partly reassured them.10
It may be better to see the issue from the Galatian point of view, in so far as we can discern it. The arrival of the Galatians in the Balkan region was fairly recent, beginning, it appears, about 400 BC. One must assume that a relatively small set of settlers arrived first, possibly preceded by raiders, or at least men who explored the region, gaining some idea of its geography and people. The eventual invasion brought in a larger Galatian force, including families, who spread widely in search of land on which to settle. The precise events in the Balkans are not known, but those in Italy, which had happened a generation earlier, may suggest the outline of the process.
The movement into Italy began, it seems, with the infiltration of individuals and small groups, some of them merchants seeking to purchase well-made Italian goods for export to the north; this took place over perhaps two centuries, before a mass movement began. As knowledge of the wealth of Italy spread, probably in the Etruscan region in particular, larger movements of whole ‘tribes’ began to arrive. The first was apparently the Insubres, from Gaul, and the Boii, from Bohemia. Other groups arrived later, and took up lands around these two. The process began c.400 BC. The earliest arrivals settled in the north and the later groups passed them to settle next along. Seven or eight distinct groups – ‘tribes’ – are attested, some arriving from Gaul, others from Germany. Between them they occupied much of the Po Valley; resistance came from the Veneti north of the Adriatic, and from the Etruscans (and eventually, the Romans) in the peninsula.11
The movement into the Balkans was partly on this pattern, but the result was different. The Keltic homeland was wide, from the Atlantic coast as far east as Bohemia, but it was, it seems, fully occupied, and overpopulation is the favoured explanation of the migrations among the ancient historians. In some cases the migrants were parts of an established tribe: one of the tribes which reached Asia Minor, the Tektosages, had branches in Bavaria and in southern Gaul, the Bavarian branch was probably the original, and the other two were probably offshoots.12 Another group, the Boii, were based originally in Bohemia, and sent out colonizing groups into Italy, and into the northern Balkans.13
The migrant groups were well organized for their move. Their purpose was to acquire land; if that land was occupied, they were most likely to establish themselves as a ruling aristocracy rather than drive out the earlier inhabitants – they wanted subjects, as well as land. The migration, however, was obviously unsettling both for the victims and for the Kelts themselves, so that one result was a restlessness among them which might emerge as raiding parties. Those who settled, however, organized themselves on the pattern of their original society, as a ‘tribe’. They may have been a cohesive offshoot of an original tribe in the homeland, as the Italian Boii evidently were, or they may have been a mixed group from several different tribes who joined together and constituted themselves into a new tribe, with a new name. Some of the names of the new tribes suggest that the migrant groups were half of one sort, with established names – Senones, Lingones, Boii – and half of the other, with new names – Ananes, Libici, Insubres. The latter group no doubt either chose a new name for themselves when they formed their union, or adopted one fastened on them by others.
The Italian tribes, when faced with opposition from the Veneti or the Romans or the Etruscans, were able to recruit Gaesatae, bands of warriors who would join the migrant groups and so reinforce their military capability.14 It was not a large step from such mercenary groups to autonomous raiding parties, and then perhaps to becoming the spearheads of the migration; that seems to have been the stage to which the Balkan migrations had developed, and when their decisive movement into the southern Balkans began.
The Kelts appear to have been in occupation of modern Hungary about the time that Philip II became the Macedonian king, c.360 BC, no doubt after preliminary infiltration.15 This was probably a migration in which the invaders established their power over the original inhabitants, though there were perhaps also larger-scale movements, as certainly happened in Italy. Once past the narrow gap between the mountains which is the site of the modern city of Bratislava, a wide low-lying country lay before the migrants, stretching as far as the Carpathian Mountains, which marked the western border of ancient Dacia, and with the Danube flowing through it. This territory appears to have become dominated by the migrating Kelts by, at latest, the mid-fourth century BC.16
This was as unsettling a movement as was that into Italy. There is no sign that contingents of Gaesatae were needed to enforce the conquest – but then we have fewer sources other than archaeology, so this cannot be taken as definitive. Certainly the flat nature of the Hungarian Plain would allow rapid conquest, especially by mounted forces. There were, no doubt, some who were keen to seek out vulnerable victims, but they were all also surely conscious of the rising power of Macedon to the south. We must assume that the conquest was quickly succeeded by a movement among the migrants to consolidate themselves into a state, and to establish their authority over their conquered subjects. The envoys of 335 BC to Alexander presuppose they represented an organized government formed by the united tribes. On the pattern of the homeland and the work of the Kelts in Italy, this political system took the form of distinct populations, described as tribes, which occupied clear territories. Of these the one which will concern this study most is that of the Scordisci, the southernmost group of those inhabiting the Hungarian Plain, established around the fort town of Singidunum, at the junction of the Danube and the Sava River. The name of any other tribes have apparently vanished, along with their territories, as a result of later subject rebellions and Roman conquest. This, of course, implies strongly that the Keltic presence was limited to achieving rule over non-Keltic subjects. Archaeology suggests that they concentrated along the Danube particularly.17
The settlement of this Scordisci group placed them on the frontier of Keltic territory, and so no doubt they will have had to keep their swords sharp, and this Keltic state will have been most concerned at the advances being made by Macedonian forces to their south, advances which had been headed northwards as the Kelts arrived.
It is, probably, therefore from the Scordisci that the recorded embassies of the Kelts came to see Alexander on the Danube, and later in Babylon in 335 and 324, and it seems clear that by that time the Scordiscian state was fully organized – that is, that the component parts of the invaders had coalesced successfully. It is not a name known from other parts of the Keltic region, so we may assume it was not originally a Keltic population from a particular original tribe; it was therefore a new tribal name adopted for the amalgam of several Keltic groups, presumably derived from Mount Scordus nearby.
This state was sufficiently alert to send out envoys who could make agreements on behalf of the whole, as with Alexander in 335. It is, however, not evident that the Scordiscian state was as yet notably powerful. If, as must also be assumed, its arrival had involved the conquest of the indigenous inhabitants, Illyrian and Thracian, the state would be as subject to rebellion and outside interferences as were Macedon’s conquests in the southern Balkans. An armed encounter with Alexander would obviously be dangerous, and we may credit the Galatian rulers with sufficient sense of self-preservation to understand that they needed to avoid a collision with such a power. The peace they had agreed at the Danube was, therefore, of the mutually hands-off type, letting both sides get on with their more immediate concerns without interference. It may seem therefore that their boastfulness to Alexander was in the nature of a bluff; that he heard it and retired to the south may have suggested to the Galatians that their bluff had succeeded.
For the Scordisci therefore the half-century from the time of their meeting with Alexander in 335 was a time of consolidation, settlement, conquest of their immediate neighbours, and increasing their population. This latter aim was accomplished in part by the immigration of further Galatians from the north, by their own natural increase, by the assimilation of conquered Illyrians and Thracians and others, and by the conversion of these new subjects into Galatians – intermarriage would help in this, as would the recruitment of these natives as foot soldiers on the promise of loot. When the invasion of Macedon came, it is noted by one historian that it was by Galatians and Thracians together, and that the Galatians were unexpectedly numerous.18 We hear of a woman, Onomaris, who is supposed to have led them in the conquest of the Autariatai, an Illyrian people, whose territory was in the western Balkans – away from any possible clash with the Macedonians in Thrace.19 There are clear indications of conquests of other Illyrians – the group called Hylli is supposed to be a mixture of Galatians and Illyrians20 – and it seems that attacks were made on the Dacians, their neighbours beyond the Carpathian Mountains.
The Dacians proved to be very resistant. There is some evidence from archaeology of the adoption of certain Keltic manufacturing techniques by the Dacians, but little sign of much Keltic settlement in the area, other than perhaps by valued craftsmen. The evidence for the Galatian presence is the findings of Keltic-type metalwork, some of which was clearly manufactured in Dacia, but this should not be expanded to imply, or suggest, or state, that Keltic settlement was extensive.21 This is an issue which will recur in this study.
The later Dacian kingdom proved to be very hostile towards the Scordiscian state, which was so damaged in the conflict that the Romans later accomplished a relatively easy conquest. The Keltic-style metalwork, so prominent in the evidence, may have been imported in trade, or it may have been made by Keltic smiths who were settled among the Dacians. In neither case can it be concluded that there were more than a few Kelts in Dacia.22
Other Galatian groups seem to have bypassed Dacia altogether, moving to the north of the Carpathians, to settle in the upper Dniestr region, where half a dozen ‘settlements’ are known. Still other groups went on to settle along the lower Dniepr Valley in the Ukraine, sending either raids or traders, or settlements into the Crimea, though it has to be said again that this is the result of archaeological discoveries of Galatian-type metalwork, which may equally have been acquired by trade. A significant concentration of finds and burials has been located in the Kyiv area. The evidence is therefore mixed: a scatter of metalwork, as in Dacia, which might be no more than the evidence of trade, but several regions where a combination of finds implies some Keltic settlement. There are Keltictype burials (with La Tène contents), widely but thinly spread, but these are individuals only and are hardly evidence of widespread settlement, unlike the concentration in the Kyiv area.23 The Dacians themselves were evidently always too tough to be tackled.
Elsewhere the Scordisci Galatians destroyed the Autariatai and sent raiding parties through the Illyrian territories from their centre in the Banat and to the Adriatic coast – whence no doubt the comments by Arrian and Strabo (from King Ptolemy’s memoirs) that they came from that region.24 The raids forced a large group of Autariatai out of their homes. They were intercepted by Kassandros, the Macedonian king, who settled them on his frontier.25 The actual result does not seem to have involved extensive conquests by the Kelts in this area, but more the reduction in power of the Kelts’ potential enemies, no doubt by being well-raided and looted, and subsequently in decline politically and militarily. Much of this is certainly assumption, though enmity between Galatians and Autariatai is certain.26 It will be noted that these raids carefully kept clear of Macedon and its unstable conquests.
The Scordiscian regime’s aggressiveness – though its existence is, I repeat, only an assumption at this point for the fourth century BC – replicated the problem which had led to the foundation of the Scordiscian state in the first place. The wider their conquests the greater their population, and this increase in population soon passed beyond the resources of the area, and probably beyond the ability of whatever central authority had been developed to exercise full control. It was part of the recent pattern of Galatian history for groups to split off and act independently. This was the source of the movement of the original population which formed the Scordisci, and of the raids which might precede the actual expansion settlement.27 There is no reason to suppose that this sequence of raids and migrations was unwelcome to the Kelts. It was no different in essence from the Greek colonial efforts of two or three centuries before, or of contemporary Roman colonizations. The traditional Galatian solution to overpopulation was to send out selected fragments of the people to establish a new home for themselves elsewhere, or simply to let the groups go. The Scordisci, by 300 BC, had been in occupation of their northern Balkan lands for up to a century, and perhaps in a few cases longer.
By this time the various original sections, Kelts and Illyrians, would seem to have become fairly well integrated, but by then any further expansion towards the west was hardly possible. Their fellow Kelts occupied all Europe from the north Balkans to the Atlantic Ocean, so expansion towards the north and west was now barred. In Italy the Romans had organized the population of the peninsula in a powerful defence against Gallic raids from Cisalpine Gaul, and had successfully defeated the most advanced Gallic tribes – the Senones were conquered in 283. There was now no possibility of expansion there, and over the next two centuries every attempt failed. Gaul, Spain, Germany and Bohemia were also all similarly out of the question because Gauls had conquered or occupied most of those lands; the Dacians were evidently as resistant to Keltic raids or conquest as the Romans. The lands to the south and south-east of the Scordisci were the only possible directions in which expansion could take place, but there also the prospect was daunting. To the south the Macedonian kingdom and the cities of Greece were well organized for defence – a siege of a city was not the preferred Galatian form of warfare – and the kingdoms which had been founded out of his empire after Alexander’s death were rich and powerful.
The precise stages of Galatian developments in the north Balkans are not known. Archaeologists have identified two elements, however, which suggest that the Danube was by no means a serious frontier line separating the Galatians from the Macedonian and Thracian territories. In that area south of the Danube, the area the Romans later called Lower Moesia, between the Danube and the Haemos Mountains, a considerable quantity of artefacts of Keltic origin have been found throughout the lands of the Triballi. This has been taken to mean that Galatian settlement took place in that territory, but this is likely to be going too far, as it is also in Dacia, for the presence of artefacts made in the Galatian style cannot be assumed to mark the presence of Galatians themselves.28 To provide a contrary example, there are finds in the Balkan region of many coins which were minted in Macedon in the reigns of Philip and Alexander, but there is no suggestion that this means a settlement of Macedonians in these areas; instead this is interpreted as a sign of tribute paid by those kings to keep the Galatians at bay, or the pay of returning mercenaries.29 Neither of these interpretations is necessary. The Galatians and Alexander had concluded a peace agreement, which both kept to; there was no need for tribute payments to keep the peace – not that there is any evidence for them. The returning soldiers, if any did return, are likely to have had many other than Macedonian coins in their backpacks – there would be Persian coins, Persian treasures, Indian items, and so on, none of which are present. For neither tribute payments nor Galatian mercenaries is there any other evidence than the finds of the Macedonian artefacts themselves.
For both there is a more plausible explanation than inventing explanations out of thin air. Trade was a vigorous practice among all the peoples of the Balkans, Galatian or not, and it is by far the most likely reason for the movement of these items in both directions. The coins moving north of the Danube out of Macedonia fit neatly with the artefacts in the south to support a balance of trade in the Galatians’ and others’ favour – supplemented, of course, by the products of the Galatians’ raids and conquests.
The Thracians, between the Haemos and the Danube, had no doubt been considerably weakened by their defeats at the hands of the Macedonians. The Triballi, for instance, are no more in evidence after about 290 BC. Extermination is not necessarily to be assumed, but repeated defeats would no doubt break up the tribe into its component sections, some groups fleeing elsewhere, some joining the raiders, some removed to the slave markets – and many dying, of course. It was the tribal structure which disappeared, not necessarily the people. Further weakening, particularly in the military sense, was produced by Alexander’s recruitment of Thracians into his army, of which several thousands of young men accompanied him in the great campaigns. There was rather more of them than those who had been killed in their war, and of these very few will ever have returned to Thrace.30 The campaign in Persia had also reduced Macedonian manpower at home. Alexander’s regent in Macedon, Antipater, had to supervise Greece as well as guard the northern frontier; further, he had to dispatch reinforcements to Alexander all over Asia almost every year.31 He was in a sense fortunate that the peace agreement with the Scordisci held. In Thrace there is evidence, largely in the form of the local manufacture of coins in the names of non-Macedonian rulers,32 that Macedonian authority was less than strong, and by no means pervasive; in particular one of the surviving members of the Odrysian royal family, Seuthes III, was developing a local power base in the northern part of the Macedonian conquests, sheltered by a section of the Haemos range.33
The Galatians meanwhile were expanding their power over their neighbours, particularly to the west in Illyria, by their domination of the Autariatai. The expansion of the Scordisci was clearly a further threat to those neighbours, and south of the Danube the divided and weakened Thracians were a tempting target. The Galatians’ victims in the period after Alexander included the Illyrians and the Dardani, as well as the Autariatai – though the Dardani perhaps only feared attack rather than suffered it, while the former two had certainly both suffered severely. It does not seem that the Thracians were directly bothered, but a branch of that people were later included in the Scordiscian state, described as those east of the Morava River.34 That is, these were Scordiscian subjects, but were still recognizably Thracian.
The break-up of Alexander’s empire proceeded slowly but inexorably, and in Macedon Antipater’s death in 319 was soon followed by the seizure of local power by his son Kassandros. He was firmly allied for the next twenty years with his neighbour, the new satrap of Thrace, Lysimachos, each reinforcing the independence of the other, which went some way to resist the local disintegration. The extent of Lysimachos’ authority in Thrace is not known precisely, and it may well have varied over time, but he was a formidable campaigner. He collided with Seuthes in a drawn battle; subsequent intermarriages of the two men with each other’s daughters seems to have persuaded both men, together with their drawn battle, to avoid further disputes.35 The whole region was clearly much disturbed. Archaeological evidence of burnt towns at Vetren, at the headwaters of the Axios River, for example, and at Pernik at the head of the Strymon Valley, is dated to this time. These two places are not far apart, and are at the western end of the great central plain of Thrace. They are also at the heads of valleys which lead into Macedon from Thrace; they were border towns of the Macedonian-controlled area, and as such were no doubt both vulnerable and obvious targets. Who burned the places, however, and exactly when, is wholly unknown; it is also uncertain who inhabited them, though it is perhaps most likely to be Macedonians.36
It is tempting always, when evidence of destruction appears in an excavation, to seek to attribute it to some known event or war or conqueror or king, and such destruction has naturally been assigned by various interpreters to all possible agencies, including, in these cases, the Galatians from the north. The attribution to any particular case cannot be accepted without better evidence than archaeology alone can supply. If it was the result of a Galatian raid, for example, this would imply there had been a considerable intrusion of Galatians into Thrace, whereas it is clear that the Galatians’ main attention at the time was devoted towards the lands to the west and south-west, into Illyria or towards their south into the lands north of the Haemos (Lower Moesia). They were not particularly numerous, and were still keeping clear of the Macedonians. There was so much fighting in the Thracian region as a whole at the end of the fourth century that any of the candidates could be responsible for the damage – or, of course, none of them.
In these circumstances it is very probable that, by moving south from the Banat area around Singidunum as well as expanding westwards, Galatian control was, by 300 BC, being imposed on the western and north-western parts of Thrace and along the Morava Valley. Seuthes’ kingdom was to the east of this, north of Lysimachos’ territory. This latter appears to have included the city of Philippopolis and the original Thracian territory conquered by Philip, but not the Haemos foothills and valleys to the north, or the Valley of the Roses in which Seuthes’ new city of Seuthopolis was being built. But west of that area and north of the Haemos Mountains it seems that at the time there was no authority of any size or power. This was exactly the sort of unorganized and divided territory of small clans and villages which had resulted from the elimination of Thracian and Triballian higher authorities. It was into such a fragmented social situation that the restless philo-migrating Galatians would choose to raid and then take over to establish control. They were not apparently seeking opposition and a fight but land, and a weak region would clearly inevitably attract them.
In 298 Kassandros, king of Macedon, evicted one band of Galatians who had reached the Haemos. Their leader was Kambaules, and his exploit is judged by Pausanias, rather implausibly, to have been the inspiration for the great Keltic invasions of Macedon two decades later, though Kambaules did not penetrate very far into Macedonian territory.37 For the moment, however, Kassandros had driven them out, or perhaps stopped them getting in. Their presence in the Haemos would suggest that there were now Galatians in some numbers in the area to the north-west, towards Singidunum. The collision, however, cannot be made to imply an intention by Kambaules to invade Macedon, though Thrace was a possible target. Kassandros’ reaction suggests that it was his own intention to enforce his power in Thrace. Kambaules therefore did not fail in an aim to invade Macedon, but the Galatians of the Banat were duly warned that Macedon, and Kassandros, were alert.
The date and place of this Galatian intrusion – it was hardly an invasion – are significant. In 301 King Antigonos I had been defeated and killed at the Battle of Ipsos in Asia Minor. Four rival kings – rivals of Antigonos and rivals of each other – had cooperated in the victory, and had promptly disagreed over the division of Antigonos’ territories. The only two of the victors whose alliance continued firm were Kassandros and Lysimachos, largely, it seems, because Kassandros had no wish to claim any of the major spoils, at least not for himself, while Lysimachos, whose individual contribution to the victory had been the greatest, took over most of Asia Minor as far as the Taurus Mountains, becoming at once the most powerful of the rival kings.38
From then on Lysimachos’ main interest lay in Asia Minor; he retained control of part of Thrace, but does not seem to have concerned himself with it for some time, and his territory was reduced in all probability to little more than the Thracian Chersonese (the Gallipoli Peninsula) and the coast of the Black Sea.39 This meant that Seuthes could flourish in independence in his new city and kingdom. Kassandros in Macedon was constantly occupied with his distracting and disturbing neighbours, in Greece and beyond his northern and western boundaries, and by 298 he was probably dying – he died the next year, of some disease assumed to have been akin to tuberculosis. In that year also he led an expeditionary force westwards to campaign in the Ionian Islands, ending with an attempt to besiege the city of Kerkyra off the Adriatic coast, which it was an old ambition of his to control; he was driven off by a rival expedition from Sicily of Agathokles, the Syracusan king – a Greek army defeating a Macedonian, to the glee of the former.40
In 298, therefore, neither Kassandros nor Lysimachos were present in the Balkan region when Kambaules began his raid, and the Macedonian king was ill as well as absent. This was precisely the sort of political situation which would attract a Galatian (or any other) adventurer such as Kambaules. It probably explains why he then made his attempt, whatever it was aimed at, though Kassandros was still sufficiently active and swift in response to thwart him. The death of Kassandros next year, and the rapid succession of his three sons in turn, the death of two of them, and the murder of their mother, was not, it seems, disturbing enough after Kambaules’ defeat to invite Galatian attention, particularly since three of the greatest warlords of the time, Lysimachos, Demetrios the son of Antigonos, and Pyrrhos the Epeirote king, were also meddling in Macedonian affairs. Any one of these could deliver as swift a defeat as Kassandros. The actual result in Macedon, by 294, was that Demetrios had taken over control, and at once he set about a massive recruitment and armament programme. Lysimachos meanwhile campaigned again in Thrace, and against the Getai north of the Danube, though, as usual, with mixed results. The Galatians were not involved. They had apparently learned, partly from Kambaules’ experience, partly from mere observation, that a powerful warlike king in Macedon was not to be challenged.
Neither Kassandros nor Lysimachos had made any attempt to contest the expansion of the Galatians from their Banat base. Blocking the intrusion of Kambaules’ raid was the only contact either king had apparently made with them, and it could be dismissed as a maverick episode. Lysimachos’ campaign in 295 or so against the Getai king Dromichaites may possibly have been a reaction to Kambaules’ raid, or to some instability connected to the continuing succession crisis in Macedon, but it may equally have been nothing of the sort. Lysimachos was sufficiently pragmatic to ignore, or accept, Demetrios’ seizure of Macedon in 294, though he did give shelter to the surviving son of Kassandros, a former king whom Demetrios had driven out, so providing himself with the possible excuse to intervene later. The main attention of all the kings during Demetrios’ reign was on his extensive and threatening military and naval preparations, which were clearly aimed at attempting to recover his father’s or Alexander’s lost kingdom. In the circumstances neither he nor Lysimachos had much time to devote to the Galatians. This is not to say that the Galatians ignored the lands to the south; quite the contrary.
By the 280s, therefore, the Kelts of the Balkans had been in place for eighty years, and at least one powerful state had developed, the Scordisci, which had established some sort of domination over a good part of the Balkan region to its south. In that time the Kelts had expanded their population, by receiving Keltic immigrants and by integrating the native populations into their state. There had therefore emerged, for the first time, a relatively stable and powerful state north of Macedon, and one which, though it had not made any overtly hostile move, except perhaps that by Kambaules, had, nevertheless, to be now regarded as a major threat to the Macedonian kingdom. Its expanding population and its political integration gave it the human resources to invade, if it chose, and its warlike methods posed a clear and major threat. There is no indication that any of the various Macedonian rulers in the 280s noticed that threat.
The half-century since Alexander had first encountered the Galatians on the Danube had brought the Macedonian kingdom and the Scordisci closer to each other, but only geographically. This was due to the political expansion of the Scordisci over the lands north of Macedon. The Macedonian kings cannot have been unaware of this, though they were generally kept busy holding on to their thrones and power in the face of competition from their Macedonian peers. The insidiousness of the Galatian advance is evident, but, apart from Kambaules’ raid, they had evidently been careful to avoid any direct challenge to Macedonian power. But they were also evidently expansionist, and extending their authority in such a way as to bring into their state considerable numbers of non-Galatians. The overall result, however, was a growth in Galatian power, and a clear diminution in Macedonian. And any Macedonian king was isolated amongst his fellows.